ssc_codegen 0.9.7__tar.gz → 0.9.9__tar.gz

{ssc_codegen-0.9.7 → ssc_codegen-0.9.9}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: ssc_codegen
-Version: 0.9.7
+Version: 0.9.9
 Summary: Python-dsl code converter to html parser for web scraping 
 Project-URL: Documentation, https://github.com/vypivshiy/selector_schema_codegen#readme
 Project-URL: Issues, https://github.com/vypivshiy/selector_schema_codegen/issues

{ssc_codegen-0.9.7 → ssc_codegen-0.9.9}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "ssc_codegen"
-version = "0.9.7"
+version = "0.9.9"
 description = "Python-dsl code converter to html parser for web scraping "
 readme = "README.md"
 requires-python = ">=3.10"

{ssc_codegen-0.9.7 → ssc_codegen-0.9.9}/ssc_codegen/__init__.py

@@ -7,7 +7,7 @@ from ssc_codegen.schema import ItemSchema, DictSchema, ListSchema, FlatListSchem
 
 setup_logger()
 
-VERSION = "0.9.7"
+VERSION = "0.9.9"
 
 
 class __MISSING(object):

{ssc_codegen-0.9.7 → ssc_codegen-0.9.9}/ssc_codegen/ast_build/utils.py

@@ -90,20 +90,59 @@ def extract_json_structs_from_module(module: ModuleType) -> list[Type[Json]]:
     ]
 
 
+def _add_sys_path(abs_path: Path) -> None:
+    # required for correct imports (eg: constants)
+    # TODO: calculate configs dirs depth
+    pathes = str(abs_path.parent), str(abs_path.parent.parent)
+    for p in pathes:
+        if p not in sys.path:
+            sys.path.append(p)
+
+
+def _is_not_dunder_obj(name: str) -> bool:
+    return not name.startswith("__") and not name.endswith("___")
+
+
+def _is_ssc_cls(name: str, cls: type) -> bool:
+    return (
+        _is_not_dunder_obj(name)
+        and hasattr(cls, "__mro__")
+        and BaseSchema in cls.__mro__
+        and not is_template_schema_cls(cls)
+    )
+
+
 def exec_module_code(path: str | Path, add_sys_path: bool = True) -> ModuleType:
     # apologize, input - real python file
     if isinstance(path, str):
         path = Path(path)
-    module = ModuleType("_")
     abs_path = path.resolve()
-    # required for correct imports (eg: constants)
-    # TODO: calculate configs dirs depth
-    if add_sys_path and str(abs_path.parent) not in sys.path:
-        sys.path.append(str(abs_path.parent))
+    module = ModuleType("_")
 
-    if add_sys_path and str(abs_path.parent.parent) not in sys.path:
-        sys.path.append(str(abs_path.parent.parent))
+    if add_sys_path:
+        _add_sys_path(abs_path)
 
     code = Path(abs_path).read_text()
     exec(code, module.__dict__)
+
+    tmp_module = module.__dict__.copy()
+    main_entypoint_schemas = {}
+    for k, v in tmp_module.items():
+        # move first schemas to end for correct order import and code generate
+        if _is_ssc_cls(k, v):
+            main_entypoint_schemas[k] = v
+            module.__dict__.pop(k)
+        # scan and import sub_schemas
+        # NOTE: context will be overrided by BaseSchema-like names
+        elif _is_not_dunder_obj(k) and isinstance(v, ModuleType):
+            for sc in extract_schemas_from_module(v):
+                if module.__dict__.get(sc.__name__):
+                    LOGGER.warning(
+                        "Schema `%s` already defined. `%s.%s` override it",
+                        sc.__name__,
+                        f"{v.__name__}.{sc.__name__}",
+                    )
+                module.__dict__[sc.__name__] = sc
+
+    module.__dict__.update(main_entypoint_schemas)
     return module

{ssc_codegen-0.9.7 → ssc_codegen-0.9.9}/ssc_codegen/converters/js_pure.py

@@ -105,7 +105,6 @@ from ssc_codegen.ast_ import (
     ExprStringMapReplace,
     ExprListStringMapReplace,
 )
-from ssc_codegen.ast_.nodes_core import ModuleImports
 from ssc_codegen.ast_.nodes_selectors import ExprMapAttrs, ExprMapAttrsAll
 from ssc_codegen.ast_.nodes_string import (
     ExprListStringUnescape,
@@ -162,8 +161,9 @@ def to_js_regexp(
     pattern: str, ignore_case: bool = False, is_global: bool = True
 ) -> str:
     """helper function for convert string pattern to js"""
-    pattern = pattern.replace("<\\/", "</")
-    pattern = pattern.replace("/", "\\/")
+    # fix backslashes translate
+    pattern = pattern.replace("/", r"\/").replace(r"\\/", r"\/")
+
     pattern = f"/{pattern}/"
     if is_global:
         pattern += "g"
@@ -187,12 +187,6 @@ def py_sequence_to_js_array(values: tuple[str, ...] | list[str]) -> str:
     return "[" + val_arr[1:-1] + "]"
 
 
-@CONVERTER(ModuleImports.kind)
-def pre_imports(_node: ModuleImports) -> str:
-    # HACK: used imports node for provide helpers scripts
-    pass
-
-
 @CONVERTER(Docstring.kind)
 def pre_docstring(node: Docstring) -> str:
     value = node.kwargs["value"]